Full text loading...
-
Efeito da Amostragem Sobre a Distribuição Espacial de Contaminantes no Solo
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 12th International Congress of the Brazilian Geophysical Society, Aug 2011, cp-264-00040
Abstract
This work aims a sampling mesh optimization to detect a possible contamination by volatile organic compounds (VOCs) in an area of storage of petroleum products, of around 20 hectares, where 2.361 soil samples were collected, 0.5 m, 1.0 m and 1.5 m deep, at each sampling point. The technique used was the ordinary kriging. The soil samples were analyzed by PID technique (Photon Ionization Detection). The strategy adopted was the analysis of krigged maps obtained from sampling with increasing reduction in the number of samples. Random and sequential sampling were adopted. Krigged maps were constructed with 100%, 75%, 50% and 25% of all available data, at each depth level. Mesh optimization was judged from three parameters: 1) the variography of each krigged map; 2) the RMS difference (Root Mean Square) between each krigged map with data reduction and its corresponding krigged map with 100% of data and; 3) the impacted area, according to each krigged map, considering each contaminated point as that with VOCs concentrations greater than 20 ppm. The sampling reduction tends to increase the RMS difference while the impacted area almost do not alters, being this effect higher for the 1,5 m depth level. Considering that the averaged range, the RMS difference and the impacted area do not change substantially with the reduction in the number of samples, until the limit of 50%, it can be recommended, for economic reasons, the reduction in the number of samples up to this limit.